Should Chatbots Pretend to Be Human?


Share on LinkedIn

In a recent post by Sandeep Raut, he writes about the role of chatbots in digital transformation.

Chatbots are built to mimic human interaction, making them seem like an actual individual existing digitally. It could live in any major chat product (Facebook Messenger, Slack, Telegram, Text Messages, etc.), powered by basic rules engine or NLP and AI.

And closes his post with this: “Going forward people will not be able to tell the difference between human and machine.”

Here’s my question: Should chatbots pretend to be human?

In other words, even if chatbots are so good they are indistinguishable from a person, should companies make it clear that users are communicating with technology?

Please add your comments below.


  1. Hi Bob, let me make a start: For the sake of transparency I would answer with a clear NO. So far it is not only about being indistinguishable but also about customer trust, which must get earned. People who would prefer a ‘real’ human on the other side and find out that they were helped by a bot are likely to feel cheated. This has a negative impact that can get avoided.

    Just my 2 ct

  2. I agree with thomas, customer trust is most important and companies should tell upfront that its a bot !!

    I just mentioned that with advanced technology it will be difficult to find the difference.

  3. Should chatbots mimic human conversation? Sure – whenever technology loses its hard, impersonal edge, people benefit.

    Should chatbots pretend to be human? No – because that’s deceptive.

    Should companies make it clear that users are communicating with technology? Yes . . . but it’s a murky area because technology is present in almost every customer interaction. It’s now become a matter of degree. I believe where there’s no human involved, full disclosure is the right thing for vendors to do.

    Any technological advance brings a mixed bag of problems and solutions. Chatbots are no exception. I am not aware of any laws requiring disclosure to consumers, so for now, the industry will have to establish its own guidelines. I believe vendors will find it in their best interest not to undermine consumer trust.

    Ask me in 100 years. I might have a different opinion.

  4. Companies should be upfront in letting their customers know, that they are chatting a bot.
    It’s the right thing to do!

  5. Sandeep also wrote about robo-advisors, and that got my attention because financial services is an industry where automated and intuitive digital responses has seen growing acceptance. If organizations are not transparent about how and where chatbots are being used with customers, and that the chatbots are offering responses only as programmed, trust could be very seriously impaired, even destroyed. The customer might then resist having anything to do with digital support interaction. Isn’t that pretty much what happened between Dave and Hal in 2001: A Space Odyssey?

  6. Wholeheartedly agree with Thomas & Andrew. Not only should a customer be aware that they are not conversing with a human, it’s going to be increasingly important that easy access to a human in all channels be improved.

    As it is, there are too many companies moving toward self-serve and AI options and making it intentionally difficult for customers to interact with a human. With eyes firmly focused on reducing costs in the short-term, they are making themselves vulnerable in the long term to competitors who will make customers feel more valued.

  7. Great question Bob.
    Not sure my vote adds much to the discussion so far, but I definitely believe companies must be transparent and make it clear when a Chatbot is doing the “talking” or rather responding.
    Today, we know when we are listening to a machine response and are OK if not a fan of it.
    As already mentioned, it’s just one other way for a company to reward their customers with honesty and openness. Trust is becoming a rare commodity these days, so any way an organisation can show their trustworthiness gets my vote.

  8. Transparency and “what’s best for the customer” are the themes of all the previous replies to your question, Bob. And that’s awesome to see unanimous among these thought leaders. Ditto!

    I welcome not being able to tell the difference between a bot and a human, in terms of quality of info and service. One of the most infuriating things as a customer today is where the automation is a huge departure from human: getting stuck in a never-ending maze of automation that does not zero-in on what I want, bores me during the lengthy list of numbers to select from so that I have to listen again, in total takes way too much of my precious time, forces me to listen to bone-crunching audio or hair-pulling repetition (there are certain tunes that while perky or famous brand jingles instantly evoke horrible experiences for me because they were used in the service queue), and requires me to repeat myself in the same contact or in follow-up contacts on the same issue.

    Being honest about automation is of course important. Designing “customer experience management” to actually improve rather than harm a customer’s view of your company is even more urgent!

  9. No!

    Chatbots that pretend to be human reek of inauthenticity and customers generally dislike it.

    On the other hand, companies do well when they give chatbots a distinct “personality.” Think Siri or Alexa, which many consumers have become quite comfortable with.

  10. Bob, thanks for the excellent question. My answer is emphatically “no”. Artificial Intelligence will always be smarter than humans, but humans have a heart; machines don’t. The strongest bond is between two people; not between a machine and a person. It’s like the “Emperor with No Clothes”. Don’t disappoint people by sounding like another person. It’s a fool’s errand. Richard

  11. Please put another check mark in the “no” column for me as well Bob, as trust and transparency are critical in all relationships. I have already heard customers sharing their frustration at being “duped” by machines posing as humans on both inbound and outbound calls! There is definitely a place for technology to support customer requests, let’s just make sure to be upfront about it!

  12. Just to clarify, we don’t use bots like this at TELUS, but our customers do share experiences they have had in other instances. They continue to express their desire for us to keep the human connection! I understand that.

  13. I can only concur with everyone who has commented on this – complete transparency in the Customer Experience is essential to start, maintain and sustain customer relationships!

  14. Ultimately it all comes down to one thing; if a customer’s issue/concern can be rectified quickly and as per their needs, it doesn’t matter if it is done by man or machine, nor do they even need to know.

    Frankly, would they even care?

  15. Frankly, I would start with “no” but remember the biggest goal is improving the customer’s experience by reducing the amount of time and effort it takes to get an issue solved. If the bot can help with that, great! And your customers will warm up to that bot quickly. That being said, it’s ok to have a little fun with the bot and align its communication style with your brand voice.

    But yeah, pretending to be something it’s not coupled with a poor customer experience is going to doom that bot pretty quickly.

  16. Hi Jeremy: my top questions are: 1) for which party do chatbots reduce the amount of time and effort to get an issue solved? With the nascent level of sophistication of chatbots now, I believe that vendors – not customers – are the principal beneficiaries. 2) How certain are you that customers will ‘warm up to’ a chatbot that can’t truly understand, doesn’t recognize sarcasm, can’t easily decipher semantic or syntactic errors, doesn’t recognize a multitude of regional idiomatic expressions, can’t sense the degree of frustration or aggravation in a customer communication, and can’t respond appropriately? I think there are way too many shaky assumptions about the efficacy of automation, and what customers really want. As service providers, if we assume that customers want efficiency more than anything else, don’t we trade off customer perceptions that they’re treated with care and respect?

    I’ll close with an anecdote about how this whole ‘efficiency’ ideal comes across. A few years ago, I went through the check out at the grocery store with a week’s worth of items. The cashier and bagger scanned a gazillion items and had the entire transaction complete in sixty seconds. In the process, they threw my stuff to the end of the belt and into the waiting bags, and couldn’t wait to throw me the receipt so they could ‘help’ the next customer. To me, it felt horrible, and I never returned to the store. I didn’t want ‘efficiency’ as much as I wanted care and respect. That’s how I feel about ‘chatbots.’ And for now, they don’t even bring efficiency – at least to me.

  17. Hi Steve, Jeremy, I guess it is not that simple. @Steve: As long as customers value being helped by humans it is not only the right thing but also extremely important to make it clear where there is a human and where there is a bot.

    @Jeremy: Measuring the improvement of customer experience solely by time doesn’t hit the mark. You mentioned it, too – ‘a bit of fun with the bot’. Andrew, too, has a good story about it, and while we are at it: Do you think that self check outs really improved the customer experience? I have my doubts.

    In summary, I too think that people will get used to bots being the main point of contact, some of us might even want that. But to get there the bots – and the companies running them – need to earn their customers trust by showing that technology is really a (vast) improvement for customer experience and not only a vehicle to drive down cost.

    And to get the trust bots must not only deliver but it must be clear that they are bots, and where the handover to a human is. Systems like Siri, Google Assistant, Alexa, Cortana will be of huge help in raising acceptance levels – or be the vehicles of its demise.

  18. Andy and Thomas, you both make great points and I don’t disagree with you. I was reflecting on self checkout a bit and why I use it? Oftentimes, I use it because it’s faster because the line in traditional checkout of 4 people long. To contrast that, we have Fred Meyer (Kroger) stores here in Oregon and they have self checkout but they also have very short lines for the traditional check stands and I often find myself gravitating to the friendly checker instead of the robot. So in that case of many stores with self checkout, “better than the alternative” still might not be a great customer experience.

    So yeah, I agree with both of you that efficiency is great for the company and sometimes the customer but we need to prioritize what’s best for the customer. And Andy, to your point, if the bot can’t understand human communication patterns, don’t pretend to be human. Like you guys, I’m really interested to watch the technology evolve.

  19. thanks for overwhelming response the question on what i wrote….yes transparency is utmost important while dealing with customers.


Please use comments to add value to the discussion. Maximum one link to an educational blog post or article. We will NOT PUBLISH brief comments like "good post," comments that mainly promote links, or comments with links to companies, products, or services.

Please enter your comment!
Please enter your name here